Using isotope-based climate proxies and model simulations to diagnose drought mechanisms in the western US

Wintertime precipitation in the western US is a significant aspect of the region's water resources because of the ability of mountain snowpack to naturally store cold season precipitation and release the moisture via snowmelt and runoff. Protracted periods of drought in the western US are typically a result of decreases in precipitation during winter months. The use of the ¹⁸O/¹⁶O composition of precipitation (δ¹⁸O_p) could prove valuable in determining mechanisms responsible for past drought because the isotopes track changes in moisture source on synoptic timescales and condensation height on subannual timescales. It has also been hypothesized that δ¹⁸O_p reflects the fraction of precipitation that falls as snow within a given season. Consequently, δ¹⁸O_p (as may be recorded in proxies such as tree cellulose) could provide climatic information that cannot be obtained from other proxies such as tree-ring widths. This study uses isotope based climate proxies from tree cellulose to better understand the cause of past declines in winter precipitation in the western US. In particular, we examine interannual variations in the ¹⁸O/¹⁶O composition of tree cellulose (δ18Oc) taken from 5 bristlecone pine trees in the White Mountains of eastern California, which capture precipitation δ¹⁸O values. Simulations of the Isotope-incorporated Global Spectral Model (IsoGSM) are used to interpret the cellulose isotope record. The simulations reveal that interannual δ¹⁸O_p variations are strongly influenced by condensation height, and secondarily impacted by moisture source. Possible dynamical mechanisms that lead to dry intervals are assessed using this interpretation of the bristlecone δ¹⁸O_c records. For example, one of the most robust features of the 5 chronologies is a local minimum in δ¹⁸O_c during 1970, which was a year prior to drought (1971-1972). Indeed, results from IsoGSM "tagging" simulations suggest that δ¹⁸O_c rose during the drought years because moisture condensed from lower in the atmosphere (where vapor has higher δ¹⁸O values). Results from IsoGSM simulations also suggest that there is no connection between δ¹⁸O_p and the fraction of precipitation that falls as snow in the White Mountains. However, further inland from the White Mountains, simulated interannual δ¹⁸O_p variations strongly correlate with the fraction of precipitation that falls as snow. The connection between snowfall rates and δ¹⁸O_p in these inland regions could prove to be a valuable tool for reconstructing drought in the western US.